Add new scan type SCAN_FOR_CONNECT to support connection manager
infrastructure.
CRs-Fixed: 2713772
Change-Id: I631f3f0324e82ef6cd8b2befbed020649c80bc4c
The function "ucfg_dfs_is_hw_pulses_allowed" should return bool and has
a compilation error
"return' with no value, in function returning non-void".
Return the bool type "false" when the dfs is a NULL object.
CRs-Fixed: 2715235
Change-Id: I8717aaedd1ac8208ce6257721d5890b82ca13a8f
The OCAC completion event 'wmi_vdev_adfs_ocac_complete_event_fixed_param'
has been updated to include center frequencies of 2 segments in MHz for
165Mhz Agile mode.
Update the corresponding Host OCAC completion data structure
'vdev_adfs_complete_status' to include 2 center frequencies to support
165MHz Agile mode.
Also FW might send a delayed OCAC completion status after the Host has
configured a new agile channel. So all the OCAC completion events should be
validated if the status is received for the current agile channel.
Update the OCAC completion processing API to include 2 center frequencies
and chwidth so that, it is possible to check if the OCAC completion status
is received on current agile channel or an obsolete agile channel.
CRs-Fixed: 2705671
Change-Id: Iab8980f8d030120d107f068d890c9694d28ee98c
Add a dispatcher API ucfg_dfs_is_agile_rcac_enabled() to determine
if RCAC is supported for the given pdev.
Change the prototype of dfs_get_rcac_enable() to store rcac
config in a bool datatype.
Also, do a structure copy of ch_params of RCAC channel in
utils_dfs_get_rcac_channel().
CRs-Fixed: 2679975
Change-Id: I4b033463fc8111144bfffd3bb7e7d2bef6568c46
Trigger Rolling CAC SM events from
- NOL expiry (EV_RCAC_START) to trigger RCAC START with a
new channel if possible and not already running.
- Agile radar (EV_ADFS_RADAR_FOUND) to restart RCAC with a
new random channel.
- vdev restart (EV_RCAC_STOP) to stop the running RCAC when
the primary channel changes. Start will be based on the
availability of the new channel.
In case of the rolling CAC feature, a channel is programmed to
the agile detector on which CAC is running continuously.
Introduce an API that finds an RCAC completed channel and return
the channel params which will be used for the next channel
change.
Also add a boolean argument in dfs_set_current_channel api to
indicate if the dfs current channel is updated in the API
"dfs_set_current_channel".
CRs-Fixed: 2674621
Change-Id: Ica54a57f131cd54e47138f1cbeef2dd0023390ed
So far the DFS algorithm detected only a train of single pulses and the
diff ts(PRI) which is the difference in time stamps between 2 consecutive
pulses, should be greater than a threshold DFS_INVALID_PRI_LIMIT of 100us.
When the diff ts is less than 100us, the radar queue gets reset. With the
updated w53 DFS specification for Japan country, new sets of duplets have
been introduced and the time interval between the 2 pulses in a single
duplet can be as low as 20us. Since, the diff ts threshold
DFS_INVALID_PRI_LIMIT is 100us, the duplets with a time interval less than
100us is not be detected.
Reduce the diff ts threshold DFS_INVALID_PRI_LIMIT to 15us for MKKN
DFS domain.
CRs-Fixed: 2656660
Change-Id: I67b8ebab80155a5bcefcaa33f5a8d75c30d2ef40
The old Japan radars that are detected in MKK4 DFS domain should also be
detected by MKKN domain.
The WARs specific to MKK4 are also applicable for MKKN.
Enable MKK4 specifi WARs for MKKN as well.
CRs-Fixed: 2673921
Change-Id: If514f7f35d7ca4fe086cc4392abdfc270eb55c4d
Dfs radar processing function dfs_process_radar_ind is called
in interrupt context. Add new API policy_mgr_get_can_skip_radar_event
to be called in interrupt context to get radar skip flag.
The policy_mgr_get_dfs_master_dynamic_enabled will acquire mutex and can't
be called in interrupt context.
Change-Id: Iffa4f56da56b991d817c36891c6f6a611f3a094e
CRs-Fixed: 2673808
An extension of preCAC feature is Rolling CAC. The distinct difference
between these features is that the FW runs the off channel CAC
timer for a "finite" time for preCAC and runs the timer for "infinite"
time on an off channel. Hence the infrastructure built for
preCAC feature is being modified to accommodate RCAC feature as well.
Following are the modifications done:
1. Add an enum to represent various off-channel CAC modes.
2. Remove the 'static' declaration of few APIS so as to re-use them.
3. Add 'dfs_rcac_ch_params' to DFS PDEV object to store the RCAC
channel params so as to use the channel params after radar detect.
4. Rename DFS APIs to match its functionality.
CRs-Fixed: 2670419
Change-Id: I0bf0d33955706941cffb4e9cf6fcebfb465a6c74
'jiffies' does not include the device sleep time.
Use qdf_get_monotonic_boottime to monitor lapse of time for nol.
Change-Id: I6e00e003b8ee3d456ac4ec62b23fdaa02d577672
CRs-Fixed: 2652692
As part of the Rolling CAC State Machine, introduce three state
enums (INIT, RUNNING and COMPLETE) and the corresponding state
events.
Introduce generic APIs such as dfs_rcac_sm_create,
dfs_rcac_sm_destroy and dfs_rcac_sm_deliver_evt for rolling
CAC State machine operations.
CRs-Fixed: 2659666
Change-Id: I528db71aa7d21dced7e47ff4f9cccfbfe94c8c21
Feature Description:
FCC/JP domains do not support PreCAC. However, we can always start
beaconning in a channel if we have completed CAC in that channel for more
than or equal to the CAC period of that channel.
If an Agile engine is available and the next channel is known, we can
start CAC in the next channel using agile, while continuing the Tx/Rx
in the current channel. And when we want to start beaconning in the
next channel after a radar detection (or after/for a user request) we can
do so without any new CAC. This allows us to jump to the new channel
without having to disrupt any ongoing data traffic. This type of
continuous CAC in the next channel while operating in the current channel
is known as Rolling CAC.
Following changes are implemented:
I)Per DFS PDEV:
1) 'dfs_agile_rcac_freq' to store agile RCAC frequency.
2) 'dfs_agile_rcac_ucfg' to enable/disable RCAC feature.
II)Per DFS PSOC:
1) 'dfs_rcac_timer' is the only RCAC timer for the device/Psoc. This will
be shared by pdev level dfs objects.
2) 'dfs_rcac_timer_running' to indicate if RCAC is running or not.
III) New APIs:
1. A dispatcher API 'ucfg_dfs_set_rcac_enable' to set rcac config.
2. A dispatcher API 'ucfg_dfs_get_rcac_enable' to get rcac_config.
3. A dispatcher API 'ucfg_dfs_set_rcac_freq' to set user configured
rcac freq.
4. A dispatcher API 'ucfg_dfs_get_rcac_freq' to get the user configured
rcac freq.
5. Rolling CAC timer init/deinit APIs.
CRs-Fixed: 2659666
Change-Id: Iae002b2ab77964fca4faf237b0d0a4e2f2afe4c2
utils_dfs_reg_update_nol_history_ch is invoked with an argument of
(void *)chan_list. However, the usage of
utils_dfs_reg_update_nol_history_ch() assumes (void *)chan_list to be a
uint8_t *ch_ieee but (void *)chan_list is a (void *) pointer of struct
dfs_channel. Similarly, the equivalent freq API
utils_dfs_reg_update_nol_history_chan_for_freq() also has a mismatch in its
usage of (void *)chan_list.
Modify the arguments of utils_dfs_reg_update_nol_history_ch() and
utils_dfs_reg_update_nol_history_chan_for_freq() API to match its usage.
Change-Id: Ic0b88606ea81bde1f7065c16dfac008b737b705c
CRs-Fixed: 2484771
Problem:
1) When operating in 36VHT160 for example, if radar is hit without
bandwidth reduction dfs_radar_add_channel_list_to_nol_for_freq will add
only DFS channels (52,56,60,64) to NOL and disable them corresponding in
regulatory channel.
2)dfs_mark_precac_nol_for_freq will add all channel non-DFS(36,40,44,48)
and DFS (52,56,60,64) to preCAC nol. 3) after NOL expiry
non-DFS(36,40,44,48) will not be removed from the preCAC tree and
therefore cannot used (as part of the full 160Mhz channel) in future to
perform agile.
While processing RADAR the dfs_process_radar_ind() prepares the
list of frequencies (freq_list) to be added to NOL. This list
also contains the non-DFS channels.
Only DFS channels will be added to the NOL in
dfs_radar_add_channel_list_to_nol_for_freq() based on the channel state.
On adding to NOL, the channel state is also changed to disabled.
The subsequent APIs that will need to process
NOL(dfs_mark_precac_nol_for_freq() for instance) will have only the
freq_list. Since the channel state is also changed for DFS channels in
freq_list, these APIs cannot differentiate a NOL channel and nonDFS
channel (as channel state is not DFS for both). Thus not adding
NOL channels to PreCAC NOL list.
Solution:
Use the NOL frequency list (nol_freq_list) that is the output
of dfs_radar_add_channel_list_to_nol_for_freq() to process NOL.
CRs-Fixed: 2655016
Change-Id: If3608405a5e8e63d93157914080dde4c01c1e0bb
Since the precac tree now supports until 165MHz, the APIs to check
if the channel has done precac or not,
dfs_is_precac_done_on_ht20_40_80_chan_for_freq() and
dfs_is_precac_done_on_ht8080_ht160_chan() needs to be updated.
Change dfs_is_precac_done_on_ht20_40_80_chan_for_freq() to
dfs_is_precac_done_on_ht20_40_80_160_165_chan_for_freq() which can
check if precac is done on a 20/40/80/160/165MHz channel.
Change dfs_is_precac_done_on_ht8080_ht160_chan() to
dfs_is_precac_done_on_ht8080_chan() which can check if precac is done
on a given 80P80MHz channel other than the restricted 80P80MHz
(165MHz) channel.
Change-Id: I90312f968547d16b2d93783138c0f870439c6c26
CRs-Fixed: 2653171
When QCA_SUPPORT_AGILE_DFS is not defined, dfs->dfs_agile_detector_id
will not be set and it will be 0, which is also a meaningful detector id.
Currently, if radar_found->detector_id is not set, or set as 0, the
detector ID is recognized as a valid agile detector which is incorrect.
Return invalid agile detector id if agile dfs is not supported.
Change-Id: I259fa6f3f9fbcb4a76567090a8f084e15ee6af93
CRs-Fixed: 2646946
MKKN domain should detect the bin5 radars of MKK4 DFS domain.
Add the bin5 radars to the MKKN domain's bin5 radar table.
Change-Id: I0d0fbe91e56f4a290a675832fe4b9974f9afbf0d
CRs-Fixed: 2645168
Instead of channel number, use channel frequency for restricted 80+80MHz
boundary check.
Change-Id: I2fa65c3b1d102acd6b64b4c6e1583d2bc29484d1
CRs-Fixed: 2645155
Currently QDF_MAX_NUM_CHAN and NUM_CHANNELS aren't aligned, this unalignment
may cause many potential OOB access. So replace QDF_MAX_NUM_CHAN with
NUM_CHANNELS to keep unified.
Change-Id: I7bf7829d776f7caf5b2afbd2c9fd0c20d608e630
CRs-Fixed: 2644073
The FW has updated the wmi command wmi_vdev_adfs_ch_cfg_cmd_fixed_param to
configure the Agile channel to enable the 165MHz channel as an Agile
channel by enabling 2 center frequencies.
For a 160MHz and 80MHz channel, the parameter center_freq1 will be the
center of the respective channels. For the 165MHz channel(restricted 80p80)
the parameter center_freq1 will be the center of the left 80MHz channel
(5690MHz) and the parameter center_freq2 will be the center of the right
80MHz channel (5775MHz)..
Update the ADFS parameters across UMAC, TARGET_IF and WMI layes to
update the Agile channel configuration command.
Change-Id: Ie480af9a6bb8dad87dd783ab06e17b449605b1c9
CRs-Fixed: 2642555
Introduce constants for building the 160MHz precac tree.
Add support to insert the 165MHz channel into the precac tree.
The precac entry is not restricted to be 80MHz bandwidth, it can be
20/40/80/160/165MHz, therefore introduce a new variable bw to
accommodate different bandwidths.
Change-Id: I57d1e2173712b54238f89079783f27ef591921d1
CRs-Fixed: 2647246
Provide support for bangradar with detector ID as one of the parameters.
Add the parameter as part of the packed arguments to be sent to FW.
Also add APIs for basic sanity check of bangradar params and packing
bangradar params inside an 32 bit unsigned integer.
CRs-Fixed: 2646549
Change-Id: Ie781bc9421b7ac0d407eb01814c9242c7f988884
In chipsets that support true 160MHz (single synthesizer),
the radar parameters received from target are different from Hawkeye
based chipsets in which two synthesizers, each 80Mhz,
produce 160Mhz signal.
The segment ID received in true 160MHz chipsets is always 0 and the
frequency offset is with respect to the center of 160MHz; unlike in
Hawkeye where segment IDs can be 0 (primary) or 1 (secondary) and
the frequency offset received is with respect to the corresponding
frequency centers (of the segment ID received).
Hence, to be able to use the same radar found algorithm as used in
Hawkeye, add a translation/mapping API that will translate new
parameters to old equivalents.
CRs-Fixed: 2633062
Change-Id: I37ad1db4ca71231119cf547abb9a8a70577cc6a4
In a future product, 160MHz mode of operation is acheived through
a single detector, unlike Hawkeye and its variants where 160MHz mode
is acheived using two 80MHz detectors.
Because of this change, the maximum segment ID is 0 in the new chipset,
whereas 1 in Hawkeye and its variants. Also the Agile detector ID
is 1 in the new chipset, 2 in Hawkeye and its variants.
In light of these changes, to identify the true 160MHz capability,
add a new API that checks the target type and returns true 160MHz
capability based on the target type. Also introduce a new dfs
variable that maintains the agile detector ID (based on the
true 160MHz capability).
In the future product, there is a support of restricted-80p80MHz
(a.k.a 165Mhz) where channels 132,136,140,144,149,153,157,161 are
available for operation as an 80p80Mhz channel.
Add a DFS API to check if this support is enabled.
CRs-Fixed: 2623964
Change-Id: If813e9d6fc649ce99c7780c04fbcb61acbd1af86
Current NOL conversion for dynamic mode switch is in such a way
that the NOL data is copied to the pdev ID entry for the targeted
mode. That is, the NOL data is split/merged before mode switch and
copied to a temporary structure before mode switch command is
sent to FW.
At this point, the target pdev's frequency range are still the old
values because of which, certain entries of the NOL data are missed.
Also in case of FW failures, the NOL entry cannot be added back
because they have been merged/split already.
To avoid the above, rearchitecture deinit and reinit NOL for
mode switch in such a way that the NOL data, as-it-is, is copied
before mode switch and is only split/merged after the frequency range
of the pdevs are properly mapped and the FW response is a success.
Change-Id: I4a073d1327ba182c40ced6089aa46d8f5f241d33
CRs-Fixed: 2632582
Per requirement, if STA+SAP is working on same 5G DFS
channel and INI g_sta_sap_scc_on_dfs_chan = 2, the
SAP DFS master is temporarily disabled to skip radar event,
and it will follow STA's channel movement. So, skip the
radar event processing based on STA+SAP concurrency policy,
otherwise the SAP would be stopped by kernel because the
current home channel is marked "disabled" by regulatory
during processing of radar event.
Change-Id: I3f851fe694c0e127665732894306e20f5d1c93f4
CRs-Fixed: 2636200
So far, only MKK4 and FCC DFS domains supported Chirp RADAR. But with new
Japan regulation, JP country which is mapped to MKKN DFS domain should
also detect the MKK4 DFS domain's Chirp RADAR.
Add Chirp RADAR detection support for MKKN DFS domain.
Change-Id: Iafd952d70726c9a9b85e73607d4a23c5022b8a46
CRs-Fixed: 2630894
When radar is detected on a channel and the radar parameters are sent to
FW, the Host Wait Timer expires even though the action status is
received from FW. At this point, the VAP state machine sends a
multivdev restart on the target channel, and is in restart progress
substate.
Since the Host Timer expires due to some reason, radar detect event is
posted to vdev state machine, that is currently in restart progress
state and, this causes a target assert and leads to kernel panic.
This issue is solved by the following 2 ways:
1) Start the host wait timer first, and then send the radar found
params to the target.
2) Check if dfs_is_host_wait_running is true, in the Host Wait Timer's
timeout, and only then call the radar_found_action_generic API.
Change-Id: I9b93c7c1822966f7d6bc00c00229849fdb5627d9
CRs-Fixed: 2604173
Currently the conversion APIs used doesn't take into account
the overloading of channel numbers between 6GHz, 5GHz and 2.4GHz
bands.
Add support to obtain the correct channel number (and auxiliary
information like band and frequency) through the new APIs that
support all three bands.
Change-Id: Ief417f14620bc2fc02d76fea3642b6fbb9615916
CRs-Fixed: 2610378
Query the external radar source and if external radars are
available use the external radars in addition to the internal
radars.
Change-Id: I9fe42ecaaf949f8e504c35ecb8cffb3e9e809fd0
CRs-Fixed: 2615278
In MCL code, QCA_DFS_NOL_OFFLOAD is no defined,
which causes the utils_is_dfs_cfreq2_ch always return
"false".
In function vdev_mgr_start_param_update
the "dfs_set_cfreq2" needs to be set correct dfs flag from
utils_is_dfs_cfreq2_ch so that vdev start wmi command will include
correct DFS flag - WMI_CHAN_FLAG_DFS_CFREQ2 for AP home channel.
Move utils_is_dfs_cfreq2_ch function out of WLAN_DFS_FULL_OFFLOAD
and QCA_DFS_NOL_OFFLOAD build flag.
Change-Id: I341db02c4db3ad3b21d8b2c115b2baffc3df3164
CRs-Fixed: 2619273
Supported dynamic HW mode switches:
DBS (full band 5G and 2G) <-> DBS_SBS (low band 5G, high band 5G and 2G)
Description of the changes:
1. NOL conversion:
a. Introduce a temporary NOL list copy structure in DFS psoc obj.
b. When mode switch is triggered:
i. Stop the NOL timers and clear the data, to avoid processing NOL
expiry during mode switch.
ii. Allocate the psoc NOL copy for the target num_radios.
iii. Store the NOL data of each radio to the target pdev ID
(pdev ID after mode switch) in the psoc NOL copy,
using a unified mux/demux API.
c. After mode switch is completed:
i. Resume NOL by re-initializing the list from the temporary psoc
copy.
ii. Free the psoc copy after mode switch is complete.
iii. Note: changes are made to support pause and resume of NOL,
increasing NOL timeout by a few milliseconds.
2. PreCAC list conversion:
a. When mode switch is triggered:
i. Stop the existing preCAC timer and send ADFS abort command to FW.
b. When mode switch is completed:
i. Unify/separate the preCAC list if the target mode is DBS/DBS_SBS
respectively, using a single API.
ii. Start ADFS again.
3. Radar detection lock:
a. While detecting radar, acquire a lock to avoid handling user triggered
mode_switch during this process. Release the lock once radar
processing is completed and CSA start is triggered.
4. Radar detection/CAC completion defer during mode switch:
a. While detecting radar or CAC completion, check if mode switch is
in progress. If yes, defer the processing and wait for mode switch to
complete before handling the events.
b. Note: Precedence is Radar over CAC, i.e., if CAC processing is waiting
and radar is received, CAC completion is no longer handled.
CRs-Fixed: 2580403
Change-Id: I506f3b569bad2e351c6f336e50f203cf5fa8b223
When FCC type 4 radar is injected at +/-30MHz separation from center
frequency in VHT 80MHz mode, incorrect pulse duration is reported in the
radar summary reports and lead to decreased radar probablity detection.
Fix the issue by capturing pulses with such characteristics and
modify them to fit within the valid phyerror pulse duration range.
Change-Id: Ic6314a372d6909448fbe4eb694c41736d1719712
CRs-Fixed: 2573339
The w53 updated pulses should be detected by the DFS algorithm.
- Add new DFS filters to MKKN filter table.
- Introduce duration margin check in confirm radar routine.
Change-Id: Icf3fecb5c6027ba827cac05dbd43a1a55463209b
CRs-Fixed: 2582300
Introduce a new DFS domain MKKN that supports new w53 Japan RADAR
pattern only.
Initialize a new empty RADAR table dfs_mkkn_radars for MKKN DFS domain.
Map The 5G regdamain MKK17 to the new DFS domain MKKN to make sure that
the new w53 Radar pulses should be detected only by JP country or
MKK17 5G regdomain.
Change-Id: Ib0f6a7e98353c9930e90703ead2342b932e491e3
CRs-Fixed: 2575347
Before mode switch command is sent to firmware, the NOL channels are
reset. But the regulatory channel structure for these corresponding
channels are still disabled. After mode switch response, when
the new umac channel list is built, these channels are still marked
as disabled in regulatory, which results in the umac channel list
not having these channels at all.
Re-enable the NOL channels in the regulatory channel list after
NOL reset.
Change-Id: Ifad8ec7a5be53e061045c068f9a6bfc313d4985c
CRs-Fixed: 2580403
Pine supports restricted 80+80 MHz only on cfreq=5690 and cfreq=5775. If AP
detects RADAR in this channel, it chooses new 80+80 MHz random channel
which is not supported by Pine because, notion of 80+80 is not present in
current channel in regulatory component. Hence, as a temporary fix we
change the invalid 80+80 MHz channel to 160 MHz or lower.
Change-Id: I749607236a1dd7b9c7aa93ff889b65adcbb4191c
CRs-Fixed: 2570057
To fix 'error: ‘num_channels’ may be used uninitialized in this
function' error, initialize num_channels to zero.
Change-Id: Ic35b606a6e075dd309844c5fca793f970a701f5e
CRs-Fixed: 2570055
Remove dummy function dfs_remove_cur_ch_from_list(). This logic is already
implemented in dfs_apply_rules().
In dfs_apply_rules()/dfs_apply_rules_for_freq(), skip all the HT20 channels
of the curent channel when DFS_RANDOM_CH_FLAG_NO_CURR_OPE_CH flag is set.
In dfs_prepare_random_channel_for_freq(), allocate correct size of memory
for leakage_adjusted_lst array.
Use DFS_80_NUM_SUB_CHANNEL macro instead of DFS_80_NUM_SUB_CHANNEL_FREQ to
choose correct 80MHz channel.
Change-Id: Ie318d667ff1b0caef3032b7aa57f0ddfb7e1f1a7
CRs-Fixed: 2570055
With the introduction of 6GHZ and replacement of IEEE channel numbers
by frequency in DFS, "leakage_adjusted_lst" is now a pointer of type
uint16_t. Change the allocation size to take care of size of the new type.
CRS-Fixed: 2569329
Change-Id: I595bc77970a2758fd6ca66e53de6c0dbfe8843bc
There may be 3 different kinds of pris in staggered pulses. If one
of them is selected as reference pri, the other two pris fail the
pre pri check, then *prev_good_timestamp can't be updated at this
time. Even the next search pri has correct pri, it can't pass the
check of dfs_check_pulses_for_delta_variance because of previous
good timestamp.
Ignore pre pri check for staggered pulses, update *prev_good_timestamp
correctly.
Change-Id: I5ce95c6c1f899f29e6b4d0d83e8171d22c31fc5c
CRs-Fixed: 2556581
As FW requires, send WMI_PDEV_DFS_PHYERR_OFFLOAD_DISABLE_CMDID when there
is no beaconing session in DFS channel for FW which supports dfs offload.
Change-Id: I73b86328be6eb132de70bd10406495fbaefcab67
CRs-Fixed: 2554083
When tgt_dfs_set_precac_state API is called, the dfs_precac_active
boolean is set to false for all pdevs. This results in preCAC
being stopped for all pdevs in DBS_SBS mode when one of the
pdev becomes inactive.
Reset dfs_precac_active flag only for the calling pdev and
set the global boolean 'precac_state_started' to false only
if all the individual pdev flags are inactive.
Change-Id: I73fcacb911a5eb9028b03aa1c86775a66f2a7fc9
CRs-Fixed: 2556734
Currently MCL and WIN both have there separate queue.h
files with the same declarations and definitions.
As part of cleanup create single queue.h in common
code and update the usage of existing queue.h to this
newly added queue.h.
Change-Id: Ie381c3553d6361cf7f8c97f14a91ed24e332b51a
CRs-fixed: 2505617
